tf.keras.layers.experimental.preprocessing.CategoryCrossing

Category crossing layer.

Inherits From: PreprocessingLayer, Layer, Module

tf.keras.layers.experimental.preprocessing.CategoryCrossing(
    depth=None, name=None, separator=None, **kwargs
)

This layer concatenates multiple categorical inputs into a single categorical output (similar to Cartesian product). The output dtype is string.

Usage:

inp_1 = ['a', 'b', 'c']
inp_2 = ['d', 'e', 'f']
layer = tf.keras.layers.experimental.preprocessing.CategoryCrossing()
layer([inp_1, inp_2])
<tf.Tensor: shape=(3, 1), dtype=string, numpy=
  array([[b'a_X_d'],
         [b'b_X_e'],
         [b'c_X_f']], dtype=object)>
 
inp_1 = ['a', 'b', 'c']
inp_2 = ['d', 'e', 'f']
layer = tf.keras.layers.experimental.preprocessing.CategoryCrossing(
   separator='-')
layer([inp_1, inp_2])
<tf.Tensor: shape=(3, 1), dtype=string, numpy=
  array([[b'a-d'],
         [b'b-e'],
         [b'c-f']], dtype=object)>

depth depth of input crossing. By default None, all inputs are crossed into one output. It can also be an int or tuple/list of ints. Passing an integer will create combinations of crossed outputs with depth up to that integer, i.e., [1, 2, ..., depth), and passing a tuple of integers will create crossed outputs with depth for the specified values in the tuple, i.e., depth=(N1, N2) will create all possible crossed outputs with depth equal to N1 or N2. Passing None means a single crossed output with all inputs. For example, with inputs a, b and c, depth=2 means the output will be [a;b;c;cross(a, b);cross(bc);cross(ca)].
separator A string added between each input being joined. Defaults to 'X'.
name Name to give to the layer.
**kwargs Keyword arguments to construct a layer.

Input shape: a list of string or int tensors or sparse tensors of shape [batch_size, d1, ..., dm]

Output shape: a single string or int tensor or sparse tensor of shape [batch_size, d1, ..., dm]

If any input is RaggedTensor, the output is RaggedTensor. Else, if any input is SparseTensor, the output is SparseTensor. Otherwise, the output is Tensor.

Example: (depth=None) If the layer receives three inputs: a=[[1], [4]], b=[[2], [5]], c=[[3], [6]] the output will be a string tensor: [[b'1_X_2_X_3'], [b'4_X_5_X_6']]

Example: (depth is an integer) With the same input above, and if depth=2, the output will be a list of 6 string tensors: [[b'1'], [b'4']] [[b'2'], [b'5']] [[b'3'], [b'6']] [[b'1_X_2'], [b'4_X_5']], [[b'2_X_3'], [b'5_X_6']], [[b'3_X_1'], [b'6_X_4']]

Example: (depth is a tuple/list of integers) With the same input above, and if depth=(2, 3) the output will be a list of 4 string tensors: [[b'1_X_2'], [b'4_X_5']], [[b'2_X_3'], [b'5_X_6']], [[b'3_X_1'], [b'6_X_4']], [[b'1_X_2_X_3'], [b'4_X_5_X_6']]

Methods

adapt

View source

adapt(
    data, reset_state=True
)

Fits the state of the preprocessing layer to the data being passed.

Arguments
data The data to train on. It can be passed either as a tf.data Dataset, or as a numpy array.
reset_state Optional argument specifying whether to clear the state of the layer at the start of the call to adapt, or whether to start from the existing state. This argument may not be relevant to all preprocessing layers: a subclass of PreprocessingLayer may choose to throw if 'reset_state' is set to False.

partial_crossing

View source

partial_crossing(
    partial_inputs, ragged_out, sparse_out
)

Gets the crossed output from a partial list/tuple of inputs.